Compare your cd playback to vinyl and post here. Please participate!

The clips are not mine BTW, they were posted by another member. I don't really see a point of analysing left & right channel tracks of 2 stereo recordings separately, since you still hear the whole clips in stereo (and if any phase cancellations are indeed present, they too will be heard)...

 

Have you ever listened to a test CD that presents completely in-phase signals on two stereo channels and, separately, completely out of phase signals on the two stereo channels? Stereophile markets a CD that includes that comparison, and they are probably not alone. In my direct experience the completely out of phase signals do not sound decreased in loudness. What happens is that when completely in phase, the resulting acoustic image sounds precisely centered between the two speakers. When completely out of phase, the resulting acoustic image has a diffuse localization--there is no precise position instruments seem to be occupying. But, and here is the thing, intensity of volume is very, very similar between the two phase tests.

I've run this. It's certainly educational (this whole aspect of phase relationships affecting spectrum plots rather blindsided me to be honest, and the effect on plots can be significant). Should you feel this stuff is worth looking into, and end up expanding your personal knowledge base, is that such a bad thing? Whatever I personally know now, well, I was hardly born with the knowledge. Learning is a life-long occupation after all.

I'm not at all trying to be your advisory. I've fumbled through all this myself. I had some plotted spectrum stereo channel versus "mid" results that made no sense to me, but once I sussed out the azimuth stuff (which I was never concerned with previously), everything I was seeing fell into place. It struck me that the plots you presented showed signs, which I had seen in my own plots, of tape head azimuth misalignment modifying the plotted results in a way that is quite different from ear/brain processing.

 

Head azimuth line up for a tape machine is a given for a mastering suite
Thats why there are tone bands at the head of the master
Summing to mono might have some very slight anomalies but these will be common to cd and LP transfers
Phase cancellations should be minimal,thats what the mono button and goniometer displays reveal in a suite
.05 milliseconds is beyond our hearing processing ability.
A stereo trace would be better ,but possibly beyond the scope of this thread

 

Change your phono cart.

What do you do for tapes from the 60s that were made without test tones?

 

The files are not mine, they are interesting.
The last cart I had was a Supex or a V15 in the early 80s.
Tapes have had line ups for ever, they may go missing in the archive
They can be tweaked by ear or now by waveform

 

Instead of them lengthy posts, just run your analysis already & reveal the truth...

 

Just because I find this subject interesting. Here are two more clips from the 'Blue' album, one from the CD and one from original LP. This time a small clip from River. Both are recorded from tape out.
I have tried to get the level more equal, but nothing is done with the files, they are both 'raw'. Personally I think both are really good, but I like the LP version a bit more.

https://dl.dropboxusercontent.com/u/39225261/R_LINE_020617-012206.wav
https://dl.dropboxusercontent.com/u/39225261/R_LINE_020617-014628.wav

 

Steve once said that tapes from the 60s do not have test tones, and have to be EQ'ed to taste.
The stereo separation on my Audio technica 150mlx is 1kHz/10kHz: 30/20dB.
Frequency response is 10Hz-30kHz
There are excellent plug-ins that will adjust azimuth after the fact,

 

The cd version has a slightly better piano, even though it is rather clangy, there is better LF and chord transient
This album ,one of my favourites, is not an audio masterpiece,it is of its time priceless
The files are close, though surface noise and clicks is evident on the LP
Thanks again for posting,getting accurate levels is difficult at home

 

All tapes have line up tones at some time, just 1Khz for reference and 10Khz for azimuth or the full Dolby NR set for accurate replay.
They may get chopped of by some dolt in the library doing a compilation, this is a great pity.
The 150 MLX looks a bit V15 ish and your tech specs need to be accurate to be meaningful Grant
10 Hz-30 KHz +/- something dB.

 

The clips are interesting and enjoyable but - they're not the same mastering, are they? (Unless I missed something.)

 

I don´t know about the mastering. I always find a difference in sound between the CD and the LP that is not really part of the media per se. CD or vinyl media could in fact sound more or less the same as I see it, apart from LPs being sometimes a bit more noisy.

 

Have you ever been advised to be careful what you wish for?

This will hardly be a less lengthy post...

Okay, I downloaded the two test tones. The first of them had been identified as the LP, with the second one the CD, and that is the case. Here are the dynamic envelopes of the clips as downloaded, the top pair is the LP.



If they don't look quite the same, it's primarily because the CD rendition has been kinda squashed into the dynamic stops at some point in its life. Sound Forge Pro reports the LP channel pair has a peak level of -0.4 dB, with an RMS average level of -16.0 dB. That's a (peak - RMS) range of 15.6 dB. Similarly, the CD channel pair has a peak level of -0.7 dB, with an RMS average level of -13.4 dB...a (peak - RMS) range of 12.7 dB. Fans of increased dynamic content might find the LP more appealing.

In the next screen capture, the CD pair has been normalized to have the same -16.0 dB RMS level as the LP pair.



The LP and CD stereo pairs should sound as loud as each other, but many ears might detect other differences in the sound.

Okay, with that background, it's time to go on to frequency plots...

As previously noted in post #60, common RMS levels are necessary for comparing frequency content plots. First up, stereo channel plots for the LP and CD after the CD RMS level was made to match the LP RMS level. The bluish region is the LP, the orangish region is the CD. That distinction will be maintained for all the following plots.



As I have learned, rumble is increasing the LP's level at the low frequency end of the spectrum. The CD's level drops very quickly above 20K Hz.

Folding the LP stereo pair into mono and doing the same for the CD pair gives the following result...this should be the Sound Forge Pro representation of what is in agreement with the plot in post #60. (Note that Sound Forge Pro uses screen real estate differently for mono plots vs. stereo plots. I've used Photoshop to compress the mono plot in the vertical direction in order to make the mono representation more visibly correspond to the stereo representation.)

To my eye at least, the characteristics of my mono plot correspond closely to the plot in post #60.



Okay, now finally on to the tape head azimuth effects that brought me into this. I am fortunate enough to own a copy of iZotope RX-3 Advanced, which provides a facility to measure average azimuth misalignment and to correct for it. Running RX-3 on the LP example, RX-3 indicated the right LP track was an average of 10.4 samples behind the left track. Since what was provided were 88.2 kHz tracks, that works out to the right LP track running about an average of 118 microseconds behind the left track. But what does that mean in terms of frequency cancellation? 118 microseconds corresponds to one full sine wave cycle at 8475 Hz, but maximum cancellation is at half that frequency...call it 4237 Hz.

RX-3 reported the CD right track to be running in advance of the left track an average of 0.8 samples, or about 9 microseconds with 88.2 kHz samples. This is pretty much down in the noise, but that works out to a primary cancellation frequency of about 55,125 Hz.

Letting RX-3 remove average azimuth mismatch for both the LP and CD, then using Sound Forge Pro once more to mix to mono and plot the frequency spectrums, here is the result. (I also ran and compared stereo plots, but they showed no difference from the stereo plots above.)



As I expected, removing the LP azimuth error lifted the plotted LP frequency response in a broad region centered at 4237 Hz compared to the previous mono plot (even though levels in that region would not sound particularly different from the previous). But I have to say there are also other changes at frequencies above that broad region I do not understand. Maybe further cancellation/reinforcement effects? I don't know. I am honestly open to suggestions concerning them.

But at any rate, I hope it can be seen that mono plots of stereo data cannot quite be counted on to tell the compete truth.

 

Here are spectrograms of the Carey tracks. The CD version flat-tops at 20k, of course, but the lp has a lot of energy from that dulcimer between 20k and 30k. (The vertical lines that go all the way to the top in both graphs are just resonances from the transfer, I believe.)

 

Wouldn't you say that even after azimuth correction you attempted the LP & CD clips are still quite significantly different spectrum-wise in some frequency areas (which was my point all along)?..

 

You may wanna set your spectrogram display range in Audacity to something under 144 dB (too much snow for my taste)...

 

No. The snow in the cd version above 20k is background noise from the recording being made from the analog outs of the amp.

Also, when a recording is made from tape, the background will be snow blue. If noise reduction has been applied, it will appear as blank like this HDTracks copy of Joni Mitchell's Edith and the Kingpin:

Not what I want to see when I buy an album. You can learn a lot from snow. Don't remove it.

 

Who's removing anything?


Your spectrograms are just painful for me to look at...

 

I'd find this subject interesting if you'd reveal what phono cartridge you're using or fill out your equipment profile.

 

That just looks like you put vasaline on the lens. Goodbye detail.

 

I agree
Azimuth error is a red herring,but it is a labour of love.
The cd has more energy from 1-10 KHz and around the 16khz field
The album is thicker around 60-100 Hz end
Any body have Mr Hoffmans version from the "master tape "?

 

What I wanted to show was that the LP and CD sound different. It must not be totally apparent what the different really is by a short listening, but to me better dynamics on the LP, and on the CD a little more treble. The analysises confirmed that (thanks guys).
I think the same pattern can be shown with many CDs/LPs compairings, as least I mostly hear the same things. The differences that we hear has not so much to do with the different medias, as that the LP and CD versions are just different, mostly.

 

The media are different, thats why people have their preferences
Many people disliked FM when it replaced AM.

 

I recently compared The Eagles Greatest Hits, The CD remaster by Ted Jensen vs. the LP with inscription in the dead wax: "Happy New Year Glynn" (Glyn Johns) and on side B: "With love from Bill". The CD was brighter overall, some instruments more up front, just loved the acoustic guitars in "Take it Easy", Tequila Sunrise", and "Peaceful Easy Feeling". It had been decades since I listened to this album all the way through, so my mind and ears were fresh for any new impressions. The wonderful acoustic renderings on the CD were immediately forgotten as I dropped the stylus on the LP. I was swept away by the LP. The vocals were more natural, lacking the stridency heard on the CD. The vocal timbre was just off on the CD. The Eagles harmonies are the primary element in their music, so the vocals have to be right for everything else to be right, even if the mix and levels differ somewhat. I am not comparing the mix nor the levels, just the sound, and my response. The CD was a bit more dynamic, so I can not criticize for any additional limiting/ compression, in fact the LP was more demure in that regard. The LP did not sound overly compressed, enough dynamics to taste, but the over the top treble of the CD, plus wider dynamic range made it less palatable. The EQ seemed to be accentuated in the lower treble region, perhaps from 2kHz to 8 kHz, quite a broad band, but the 2K to 4K range was just too much IMO.

I have a Dual 701 turntable on the shelf, and a 1229, as well as an original AR XA, all modest turntables and a rather nice Shure "Premier" on the 701 (which is an M-95 with hyper-elliptical stylus) But I only have my little Technics SL-D205 direct drive connected at my desk, and listened through a Pioneer SX-780 receiver I had just refurbished (so it performed like new and was soon sold to very happy new owner) and through vintage Sennheiser HD-420 headphones. The cartridge is a Shure M-91ED with a better generic stylus that tracks very nicely. The M-91 is a rather thick sounding cartridge with an extended, sweet top end. So taking into consideration the cartridge warmth, the LP should sound closer to the CD on a more neutral cartridge, but I doubt the treble would ever be so harsh as the CD was. My CD player connected at the time was an NAD T-550 DVD player with a Burr Brown 24/96 chip.

The high frequency extension of the LP was immediately apparent, overall a sweeter presentation without fatigue. I listened to the CD first, followed by the LP, then a second listen to a few tracks on the CD. I will admit, the acoustic guitar treatment on the CD added new life and excitement to the tracks I mentioned above, that lacked by comparison on the LP. The Jensen remaster is a different mix and I believe some tweaking of the EQ.
Steve VK

 

Ive just been listening to the Eagles Greatest Hits CD and others too
I think I prefer Glyn Johns leafy Barnes sound to Bills Californication
Both are v good, just either side of a large pond.
I dont think Glyn got on with the chaps though.
As for spectoral content ,I like detail, and it depends very much on your cans/LS of choice for a pleasing finish
The HD 420 are indeed vintage,Sennys second attempt at openback and the hifi market.
My scanspeak tweeters and midbass drivers are very neutral, these cds don't sound bright to me on my HD 650s or LS
What do you think the session tapes sounded like?