WM-D3 recording - balance issue

Discussion in 'Tech talk' started by dom69, Nov 30, 2023.

  1. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Hi everyone,

    I'm progressing with my WM-D3 repair, now everything has been put back together and I could try a first recording.

    I have to say that playback quality is not really good on this unit (am a bit disappointed I must say) so I play it back on my WM-DD 33 which has far better sound quality. Indeed it's like day and night, recording is really good except that I have a noticeable balance issue between left and right channels, the left one being louder than the right one.

    Is there a simple way to adjust this without oscilloscopes or complex hardware (that I don't own)? I'm not looking for perfect as-per-specs measurement, simply aiming to fix my balance issue.

    Thanks in advance for any piece of advice!
     
    Last edited: Dec 1, 2023
  2. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    Yes, the simple way is to simply adjust it by ear. Rotate the trimpot until you can no longer hear the imbalance. Rotating counterclockwise increases the level, clockwise decreases it.

    Of course this method has a couple of issues (which are not as minor as they might look at first glance):
    1. You don't know which channel you need to adjust unless you do a measurement of the record levels on the line out and compare REC with PB
    Generally speaking, it's likely the lower one needs adjustment, but there's no way to know for sure without measuring.
    However a measurement would not be very releavant if not the PB levels are adjusted first.
    2. You don't know how much the PB levels are contributing to this issue. Those are usually off, even though by listening the difference may not be obvious.
    By adjusting only the rec levels you compensate in recording for a problem that is likely to be partially caused by an imbalance in the playback (Dolby) levels.
    3. The bias adjustment (which is L/R independent) also influences the record levels.

    So, if you want just a rough adjustment by ear, do that. However do keep in mind the frequency response on recording is likely to be off (even if it seems to sound ok) and Dolby encoding won't work properly.
    In my opinion, the most important adjustments by far are the playback (Dolby) levels because these will influence everything else. Unfortuately this adjustment requires a calibration tape.

    EDIT: Also, how do we know it's an adjustment problem ? It can be something else.
    You need to check that when feeding it a balanced signal you do get a balanced signal at the rec amplifiers and they don't get attenuated somewhere in between the line in and rec amps.

    As for DD33, I agree that one has significantly better sound. But even the D3 is not that bad when everything is properly adjusted on it.
    An upgrade I recommend on the WM-D3 is replacing the 2 x 10uF SMD capacitors in the head preamp with 2 x 47uF which can be found in this kit: https://fixyouraudio.com/product/sony-wm-dd3-capacitor-upgrade-kit/.
    Additionally you can also replace the 2 output coupling capacitors (220uF --> 560uF), which can also be found in that kit. This upgrade will significantly improve the dull sub-bass response this unit has.
    Note the rest of the capacitors (apart from the 4 mentioned) in the kit won't be needed for this model.
     
    Last edited: Dec 3, 2023
  3. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Hello Valentin,

    Thanks for your response! I've been doing some research, and came across the following tutorials (yours actually :)), which I will follow:
    https://stereo2go.com/forums/threads/sony-wm-d3-restoration-tutorial.6884/
    https://stereo2go.com/forums/threads/sony-wm-d6-d6c-calibration-tutorial.7164/

    I've read and re-read them, and I believe I get most of the idea, but still have some questions and points that need clarification. I must say I have absolutely zero background in electronics, electricity etc... so I'm really learning on the fly (same for soldering) - and am quite worried when attempting anything :p

    Sorry in advance if those are dumb questions, please bear with a total neophyte.

    Parts I will need:

    Dolby Level Tape
    : the link you mention does not exist anymore sadly, I've found an alternative option (https://www.ebay.fr/itm/176013927759) which is a bit expensive for the one cassette I will be using - do you know of another (less expensive) source?
    I have also found an option for a P-4-L300 tape, but you don't recommend it, and it's not that much cheaper anyways...

    TrueRMS Multimeter: I'll be using this, you mention BNC to Banana adaptor but I'm assuming I can build a RCA to Banana, will be the same, correct? (possibily dumb question, I don't know anything about impedances and I guess there is no difference in that respect between the two). I would then connect my standard mini-jack to RCA cable to my home-made RCA to Banana and plug it to the multimeter (with the resistor).

    Azimuth adjustment:

    I won't touch anything, treeble response seems OK, again not the best sound out there but I guess nothing to do with my problem.

    Playback Level Adjustment:
    I understand this also means Dolby Level adjustment. The Service Manual specifies to make this adjustment with the DOLBY NR switched off... Is this correct? I will set the tape type depending on the actual Dolby/P-4-L300 test tape type, but am not clear about this other one.

    Record Level Adjustment:

    REC LEVEL: The service manual specifies that this adjustment should be made at "Standard Record" REC LEVEL (described in page 11). Do I correctly understand that I simply should set the record level so that the internal VU-meter lights the -5 dB diode (to match the output level of 0.44V mentionned) when receiving the signal? I will be using something like this.

    INPUT/OUTPUT monitoring: here I will monitor the signal at the headphone output, that is clear, but:
    1. do I need to have the 47k resistors there as well, or is it only for when using the LINE OUT output?
    2. what volume should I set? does it matter at all? or, since we are simply going by comparing input/output signal, it doesn't make any difference as long as it's kept fixed across all the testing...?
    3. I should then adjust the recording trimpots to ensure output level matches input one for both channels (not accounting for potential imbalances between the two)

    Bias Adjustment:

    Is this one absolutely needed, or can I stop when the previous one is OK and I feel I have overall good, balanced sound? Or, is it mandatory to recheck this as soon as you meddle with recording levels? I'd rather avoid soldering if I can :p

    Thanks again for all the help!!!!!
     
    Last edited: Dec 5, 2023
  4. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    About the calibration tape, the one I recommended initially in the old thread is unfortunately no longer made.
    The one in your link is a good alternative. It is more expensive because it's a set of tapes for wow&flutter, azimuth, Dolby levels, crosstalk, etc.
    If the newly made P-4-L300 are good, haven't used one so can't say if they're good or not. The old ones I certainly don't recommend.
    If you find other alternatives please share. At the moment I can't recommend anything else, but if I find something will share.

    Playback levels and Dolby levels are one and the same thing.
    Dolby NR needs to be OFF for the adjustment because the aim is to create a correlation between the magnetic fluxivity and the tape and the signal levels after the preamp, not to adjust for Dolby response.
    Dolby NR is level-sensitive, meaning it acts differently at different signal levels. See here how the effect is much stronger at low levels: https://en.wikipedia.org/wiki/File:PlotDolbyCEncodeCurves.tif
    EQ (120/70us) should be set accordingly to the tape that is used. All Dolby level tapes I have seen are TYPE I, but if the tape would be TYPE II set the EQ switch accordingly.
    The line out does need to be loaded with 47k for the playback levels adjustment. For this adjustment you do need to use the line out, headphone out won't do.

    Adaptors used for multimeter (or other measurement device you use) don't really matter. We're not talking about RF stuff here. Use/make any adaptor or multiple adaptors that fit your particular situation.

    Azimuth: it should be checked and adjusted if needed (it is off on many DDs). The fact it seems ok doesn't necessarily mean it is.

    Record level
    : ignore what the service manual is saying as their procedure is very different than mine (it's more complicated and inaccuarate). This is something I stated in the original thread as well.
    The level at which you do the adjustment doesn't really matter, what matters is that you don't saturate the tape. Hence why it's best to use a level of -5dB or lower (usually I use -10dB) on the VU-meter.
    Volume level obviously does not matter, just set it to a reasonable value (like 3-4) so you're not close to the maximum where the amplifier will clip.
    In regard to loading the output, while it's not critical I would still recommend doing it in order to keep the amplifier's output level as stable as possible.

    You need to adjust so what you monitor when recording (we'll call this the input) is the same as what you will get when rewinding and playing back the recording (we'll call this the output).
    And yes there will be an imbalance between the channels, don't account for that. However if the imbalance is very high, that is an indication there is something else wrong.
    You can account for the imbalance, problem with that is: once you do it the rec levels after the Dolby encoder won't be right anymore and Dolby encoding will be off (remember Dolby is level sensitive).

    Bias: It is absolutely needed ? Depends on 2 things:
    1. Will you record using Dolby NR ? If so, it is needed as otherwise recorded tapes will sound muffled when engaging the Dolby NR on playback.
    2. How many different tapes you intend to use ? Bias settings highly depends on the tape forumulation, so it will only be perfect on the one tape the unit has been calibrated on.
    So if you use many different tape brands and models, the result will be far from ideal anyway, in which case the bias setting will matter even less.

    So, as stated before: if you ain't going to use Dolby and don't care so much about having a flat frequency response on recording, doing the simple adjustment by ear will be enough (assuming there is no problem between line in and rec amps).
     
  5. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Thanks Valentin, I've order my small material (resistor, plugs) and the Dolby calibration tapes set and will update when I have some progress.
     
    Last edited: Dec 6, 2023
    Valentin likes this.
  6. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Hello there!

    I finally received everything ordered (Dolby tape mentionned above, a TrueRMS multimeter, and small parts to build my cabling rig) and could perform the adjustments.

    I started by building my connectors (thanks to your post on the WM-D6 calibration tutorial):
    Rig.jpeg

    I could then start measuring my PB levels, and discovered indeed a difference between the 2 channels. Eventually I managed to set PB levels so that I read an average (was fluctuating but I guess that's normal) of 340 mV for each channel with the Dolby calibration tape (Dolby mode OFF).

    Then, I went for the recording levels, and actually had to adjust them quite significantly (i.e. with an input of 50.85 mV for right channel I started with an output of 37.8 mV, and with an input of 57.5 mV for left channel I started with an output of 47 mV). I adjusted trimpots so as to have output level matching input one for both channels, not accounting for the imbalance.

    I then checked PB levels again and had to micro-adjust to reset them to 340 mV for each channel (only by 1-2 mV approximately), and went for a recording... Tadam!!!... Pretty good to my ears, seems to be spot on - I guess my brain has to forget the initial imbalance, and that I'll need a few recordings from various sources to fully feel relieved, but I believe my problem is now fixed :)

    Last questions:
    • Is the re-adjustment of the PB levels after recording levels something usual? or should I have left it alone (I can go through recording levels adjustments once more now that I know how to do, if I should not have touched PB levels afterwards)?
    • Is the 13% difference/imbalance between left/right channels on recording levels considered within "normal discrepancy" range? And why doesn't an imbalance actually matter??
    • The recording levels seemed to be pretty off between INPUT/OUTPUT for each channel, is this usual? this is supposed to be a high-end unit, weren't they properly calibrated in factory? Mine seems to have fallen at some point (I fixed a bump on the body) - could this explain? (but I doubt a simple fall would rotate the trimpots..?)

    Thanks again for all the excellent advice - you guys rock!
     
    Last edited: Dec 23, 2023
    Emiel and Valentin like this.
  7. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    Glad you got everything done !

    In regard to the order of the adjustments, the PB levels need to be done first because these influence the other 2. The REC adjustments, however, do NOT influence PB adjustments.
    The only situations where it might be necessary to re-adjust the PB levels are:
    - if you replace the head, do a head lapping or if you adjust the azimuth;
    - if you replace the 2 x 10uF capacitors in the preamp feedback circuit with larger values (47uF) to improve the bass response;

    1-2mV variations on the PB levels are irrelevant (it's normal to have some variation, it's analog tape), the tolerance on Dolby B is way higher than that.

    About the imbalance in REC levels, you have L=57mV and R=51mV. It's a 6mV imbalance, that's a little below 1dB something I would argue isn't really audible.
    It is normal, because stereo potentiometers have a bit of imbalance. And in this measurement (done on headphones out) there are 2 pots influecning it: REC and volume one.

    It's also normal for the REC levels to be off because:
    - PB levels are off. These are off from the factory (in most cases) as the calibration tapes they used at the time were off;
    - the levels (and bias as well) will vary a lot with tape used, so if you're not using the exact same tape used at factory, you should expect them to be off even if PB levels were to be perfect;
    - the original adjustment method (presented in service manual) is inaccurate.
    This original method implies generating a very precise input voltage into the MIC in and then adjust for an exact output level when played back.
    Main issue with this method is it doesn't account for any L/R imbalance nor does it account for any mismatches in the gain of the amplifiers.

    About dropping the unit, while of course it's not impossible it can affect the adjustment, don't think that is actually the case.
    I've adjusted many D6Cs and D3s and can tell these adjustements are almost always off.
     
    Last edited: Dec 24, 2023
  8. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Thanks again for all the advices Valentin!
     
  9. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Hello again, and happy new year to everyone!

    After a few recordings and listening sessions, although it's definitely much better than my original imbalance issue and close to perfection, I still have the feeling that right channel is very slightly lower than left one when listening to recorded tapes. Voices are not perfectly centered, I can still feel a very slight imbalance.

    I initially adjusted PB & record levels, but did not touch anything on Bias.

    Questions:
    • could this remaining imbalance be linked to bias? Since it can be adjusted for each channel, can this be a possible cause?
    • what else could explain this? is there any other component in the chain that could explain this - and how can I check it?
    • could this still be linked to record levels and this 1 dB difference?
    Cheers!
     
  10. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    I would do the bias adjustment as well just to know it's done, however a bias "imbalance" would cause a difference in frequency response (like one channel would sound a bit muffled) not a level imbalance.
    What I can't explain myself is how the levels measure good on the meter, but despite this there is audible imbalance. Maybe what you have is a channel that rolls off much quicker than the other.

    You need a 10kHz test signal for this adjustment and make the input equal to the output (use -10dB levels). Given you have only discrete steps, choose the value that's closest to the input, but lower.
    For example, if you have 100mV input and in configuration 1 you have 90mV output, while in configuration 2 you have 110mV output, choose configuration 1 with 90mV.
    Note that for increasing the level you need to decrease the total capacitance. You have large values and a 2.2pF small capacitor for fine tuning.

    There are many things that can cause an imbalance, but it's not a good idea to start guessing. First do the bias, then re-do the rec levels and we'll take it from there.
    If after all these adjustments are done the problem is still there, I can suggest other things to test.
    So, do the bias adjustment and see how big of a difference in 10kHz response you have to begin with.

    I would argue 1dB of imablance isn't really audible, it's higher than that given the fact it's obviously audible.
    The levels don't have to be 100% perfect to have no audible imbalance.
    From your description it seems the imbalance has just been lowered to more acceptable levels, but the main problem you started with is still there.
     
    Last edited: Jan 10, 2024
  11. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Hello,

    I finally went through the bias adjustment procedure. I've tried a couple of different settings for the left channel, and when finally trying again the original one (15+18+22 pF) it actually proved to be the one closest (but lower) to match - strange as my first readings before touching anything showed a much bigger difference between input and output - could the unsoldering/resoldering have changed anything? Right channel was close enough, so didn't touch it.

    Then I readjusted record levels, and went through another music recording & listening test. Unfortunately issue is still there, zero improvement.

    I have to add something that may help better understand the problem. When I first listened to cassettes on this WM-D3 unit, I immediately noticed that playback quality totally sucks, so all my listening tests (not the measures) were made on my other WM-DD 33. I've never bothered to try again to actually listen what I record on this same WM-D3 unit.

    My procedure has been the following until now:
    • make all measures, adjustments etc... strictly following your tutorials, on the WM-D3 (using Line In, Line Out and Phones outputs)
    • for my listening tests, I record music on the WM-D3, and play it back on the WM-DD 33
    Today I've listened to my recording on the WM-D3 and realize that:
    • quality definitely and totally sucks - frequency spectrum is completely different, I totally lack in low and medium frequencies
    • there is a huge echo, and imbalance is much more noticeable
    • it's really like I would hear the music from a distant place in a cathedral, and it would be coming from my left side
    So what I can say is that playing back on my WM-DD33 is close to perfect (all frequency spectrum is there) except this slight imbalance (left channel very slightly louder than right one), but when I listen on my WM-D3 it's like total crap and completely unbalanced.

    EDIT: Something just crossed my mind... PB levels are measured through Line Out, but record levels and bias are adjusted by Phone output... If there is anything wrong with the circuit (by lack of better word in my vocabulary) of the Phone output (as seems to suggest my disastrous listening experience) could it explain the issue?

    Any ideas on where to look next? :(

    Thanks!
     
    Last edited: Jan 20, 2024
  12. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    To me it looks like you have an azimuth problem, so I would go in that direction. First adjust by listening to see if adjusting it does actually improve the sound.
    First you need to have good playback, otherwise all the adjustments will be off.

    In regard to the bias adjustment, desoldering and re-soldering itself don't see how that would influence the adjustment.
    However if there was some residue of flux or other substances between pads that has been cleaned that could explain it as residue can create parasitic capacitance.
    And hence the capacitances in that circuit are small, adding another 5pF of parasitic capacitance can change the bias levels.

    As for the headphone amplifier being the problem, of course we can't exclude that possibility although it's unlikely.
    However, there's a very simple way to test if that is the case: use an external amp to play through the line output and see if the problem is still there.
    Even without a headphone amplifier, you can still plug the headphone directly into the line out and listen (volume will be very low).
     
  13. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Thanks Valentin.
    Reading other articles or from your tutorial, I understand that azimuth would impact high frequencies, not balance. Also, how could a tape recorded on one unit play well on another, if azimuth is dead on the recording one?
    I don't have oscilloscope to adjust this and am a bit worried to fiddle around with it and end up completely messed up. Maybe I'm too worried, and it actually is fairly easy to adjust it by ear - I have an azimuth test tape from the set I bought for the Dolby tape, is it just about playing/listening to it and adjust screw until I hear a maximized and balanced output?
    Indeed I cleaned pads with alcohol in between soldering trials so that is probably the reason.

    Before:
    Before.jpeg
    After:
    After.jpeg
    I will start by checking this - what symptoms should I look for to confirm if there actually is an azimuth issue? I would assume that since PB levels were adjusted through line output, there shouldn't be any balance issue...

    EDIT: I have listened to my recorded music through Line output on the WM-D3, and what I can hear is definitely much, much better than via Phone output, and close to what I have on my WM-DD 33 (with the same slight balance issue), but with a lack in high frequencies.
    I also have checked the azimuth adjustment screw, and the lock paint seems to still be there.
     
    Last edited: Jan 21, 2024
  14. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    Azimuth can create a slight imbalance together with high frequency loss, it depends how far off the adjustment is.
    In regard to the adjustment, I think you shouldn't be that worried indeed as you can easily re-adjust it very close just by listening.

    For reference, I don't recommend the oscilloscope adjustment because it doesn't show the whole picture.
    Observing the Lissajous figures on scope doesn't guarantee a perfect adjustment as there are 3 correct ones (in phase, 45 degrees and 90 degrees),
    but only one of them also has the maximum levels. And you will need to switch back and forth between normal mode and XY mode while the signals are fluctuating. Not ideal.
    Rather, measuring the frequency response with a white noise test tape (you can do that with free software) and adjusting for the flattest response is the way I prefer to do it.
    For this you will need to record your own test tape on a deck that can do calibration on the specific tape formulation you're using (either manual or automatic).

    If the lock paint is still there, I would leave it as is. But it's best to double check with a screwdriver if that screw is indeed locked or turns very easily.
    I've seen situations where the screw seemed like it was locked, but when I put a screwdriver on it, I realised it was actually loose.
    Missed this particular piece of information. So from my understanding the problem manifests just on playback and just on the headphones out ? Is this correct ?
    Think the simplest way to rule this out is to simply look at the head relative to the head bridge and carriage. Is it parallel to the head bridge/mechanism chassis ?
    Usually, when I do head lapping (which requires the head to be removed) can do a 90% good azimuth adjustment just by visually aligning it to be parallel with the bridge.
    Of course after this I do the adjustment via measurement and then lock the screw, but the idea is the visual inspection can tell a lot.
    Ok, so let's clarify: you get much better sound via the line out, but the imbalance is still there ?
    This imbalance is also present when you play the tape on the DD33 or just on the D3 ?
    When you listen via line out, you have a lack of high frequencies ?

    A recording of the sound would be ideal if possible, both via line out and headphones out. I would also be interested in hearing both a tape recorded by the D3 itself and a known good pre-recorded one.
     
  15. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    For the azimuth adjustment screw lock paint I will double-check as you suggest with a screwdriver. From visual inspection the head seemed to be parallel to the head bridge/mechanism chassis, but I will also double-check that point. Thanks for confirming I can do the adjustment by ear if necessary - unfortunately I don't have any other recording device so can not create the white noise tape on a known, working deck...

    Therefore, if the screw is loose, I will go and adjust by ear based on my azimuth test tape (10 kHz -10dB), searching for the maximum, balanced output, through Line Out.
    The problem is much bigger via headphones out. Via line out, it's only slight imbalance and lack of high frequencies (see below). So there is something wrong somewhere (imbalance and high frequencies), and depending on the output it jumps to your face or is more subtle to detect.
    Yes, all correct.
    Recording music from my iPhone to the WM-D3 and playing it back on the WM-D3:
    • via line out, I have relatively good sound, with the slight imbalance audible, and lack of high frequencies - the rest of the spectrum being OK-ish. It would still be below WM-DD 33 sound quality but somehow acceptable (in any case I wouldn't use the WM-D3 to listen, just to record, due to the difference in playback quality).
    • via headphones out, crap sound, strong imbalance, strong lack in low and medium frequencies.
    Playing it on the WM-DD 33 is showing the same imbalance than via line out on WM-D3, but with all the frequencies spectrum (incl. high ones), matching the iPhone source sound quality (except for the slight imbalance).
    I will try to do that but can take some time, I'm not really equipped to record on my laptop.
     
    Last edited: Jan 22, 2024
  16. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    In regard to the recording, you can get an external USB sound card with line in, like this one: https://www.ebay.com/itm/171800234043
    It's enough for the job and will be useful for future repairs: measuring wow&flutter, frequency response and making recordings from such devices.

    I mostly understood the problem, however it's still unclear to me if the WM-D3 can play a pre-recorded tape (if you have any) properly or not. Have the same question in my mind about the DD33.
    From your description, it seems like you have problems in the playback circuits that create a muffled sound and imbalance.

    The fact it seems better via the line out, not really sure based on your description if that just seems to be the case because the volume is much lower or the sound really is different throught the line out.
    What you can do to confirm this is lower the volume on the headphones out enough that you get a volume match with the line out.
    Then compare, plug the headphones in one jack the the other, playing the same tape. Notice if there's an obvious difference. Would expect a slight difference due to the fact line out is high impedance and won't be able to drive headphones as well.

    Did you have any battery leakeage on the PCB ?

    What I'm really puzzled by is how the levels measure perfect, but still there is a very audible imbalance. Maybe it's not imbalance, but crosstalk between channels.
    Maybe the frequency response of L and R channels is very different, sounding like an imbalance. Without listening to it, it's not obvious what the issue is, hence why I asked for a recording.
     
  17. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Thanks for the tip, I've ordered one :)

    I've tried a pre-recorded tape on the WM-D3 and I believe that it plays well through Line Out, however via headphones out I have the same issues (imbalance and frequencies). I will try a few recordings as soon as I receive the sound card and share them with you.
    I'm using a noise cancellation headphone (Bose QC15) and headphone volume is set at 4, matching the line out one. It's really sound quality, not a difference in volume, as I'm switching between the 2 outputs while the tape is playing. I'll try to send the recording.
    None that I could see, but I'll check again.
    I should be able to send samples by mid-next week, hopefully I will have the card by then.

    Thanks again for all your help!
     
    Valentin likes this.
  18. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    Hi, having received my external sound card I've been able to make a couple of tests a few minutes ago. I'm not finished but I'll share my first results and findings.

    The 3 attached audio files are from a known good tape I just bought on Amazon (Guardians of the Galaxy soundtrack). It is a Type 1 tape, with Dolby NR set to ON on both the WM-D3 and WM-DD 33 units. I've tried to capture with the same volume generated so as to have a good comparison basis. I've captured from the WM-DD 33, and then from WM-D3 from both the Line and Phone outputs.

    When listening to the captures before sharing them, I was struck by the fact that I didn't get that nasty difference between Line and Phone outputs on the WM-D3. When listening again directly from the WM-D3 with my headphones plugged in, I still had my crappy sound, until I realised something that completely escaped me before: my headphones have a TRRS connector (there is a microphone in the cable), whereas my jack-to-jack cable for capturing through the external sound card is a standard TRS (no mic) :confused:

    So the strongest difference between Line Out and Phone Out on the WM-D3 was due to a wrong contact in the connector plug (confirmed by slightly pulling out the headphone connector out from the plug, which fixes the issue when the connections are properly done) ==> 1 problem "solved" :rolleyes: Strange that I don't have the same with the WM-DD 33......

    Now, even with a pre-recorded tape, I can still hear some imbalance (or what sounds to me like some) on the WM-D3 which I don't have on the WM-DD 33. Typically, listen to the drum cymbal on the right channel around 20 sec and the difference between my WM-DD 33 capture and the WM-D3 ones - huge difference.

    What I can't really explain is that the difference I hear via the headphone between Line Out and Phone Out on the WM-D3 (even when cable is slighlty retracted to avoid the TRRS effect issue) is not really audible on the captures. Via my headphones, Line out is really bassy and lacks strongly in medium and high frequencies compared to Phone output, but on the captures it's similar - no idea why.

    Tomorrow I'll record the same song from my iPhone on the WM-D3, and will share the same captures afterwards (from the WM-DD 33, and then from the WM-D3 through Line and Phone outputs).

    EDIT: I see the capture files are reduced in sized during upload process, I hope you can still hear the differences.
     

    Attached Files:

    Valentin and Emiel like this.
  19. Valentin

    Valentin Well-Known Member

    Messages:
    1,618
    Trophy Points:
    113
    Location:
    Europe
    I wanted to ask exactly that before my previous message, but after reading you used a pair of wireless headphones I assumed the mics are in the cups and they're probably using a normal 3 pole jack, so I simply skipped the question.
    This issue with 4 pole jacks I have seen on some walkmans and some portable CD players, although not all of them have this issue. On vintage devices it's prefferable to use a 3 pole cable, eventually an aftermarket one.
    I incline to blame it on the cable. If you have a different pair of headphones or IEMs, I do recommend trying them just to eliminate this as a potential problem.

    I listened to the files, can't hear an obvious difference between line out and headphones out, the problem seems to be there on both recordings and it is an imbalance.
    Importing the file into Audacity and doing a spectrum plot reveals there is no significant difference in frequency response between the L and R channel, but there is a strong imbalance (about 5dB) - see second picture attached.
    What is very curious to me is how this imbalance doesn't show in the PB levels measurement... Try to do the measurement both with and without 47k loading on line out, see what kind of difference do you get.
    Please post a picture of the PB level trimpots, I'm curious in what position are those relative to each other.
     

    Attached Files:

  20. dom69

    dom69 Member

    Messages:
    78
    Trophy Points:
    18
    Location:
    Switzerland
    I'll try to find another pair of headphones and check if I get the same differences, however after writing you yesterday evening I managed to find another cable (without the mic) for my headphones (the QC15 is actually only cable not wireless, and came with 2 cables that you can switch) - and same result and differences between Line Out and Phones Out. Weird.

    OK I'll do again the PB levels measures (with and without the 47k load) and share. In the meantime, I've attached the picture of the trimpots:

    PB level trimpots.jpg
     

Share This Page