Assisi Posted September 15, 2020 Posted September 15, 2020 On 10/09/2020 at 6:31 PM, Assisi said: I do not have a 12V LPS. I will consider one later on. @dayday, I did not wait very long on the matter of considering a 12v LPS for the Rhein Z1. Today I received a Weiss PSU101 in the mail. A very serious combination. Just lovely. I expect things to settle into more improvement over the next few days. John
dbastin Posted September 15, 2020 Author Posted September 15, 2020 (edited) 1 hour ago, Assisi said: I received a Weiss PSU101 Hèy John, it would be great to hear about how that Weiss LPS performs. Can I suggest to start another thread so this one doesnt drift much off topic. Edited September 15, 2020 by dbastin
Assisi Posted September 22, 2020 Posted September 22, 2020 (edited) On 15/09/2020 at 11:22 PM, dbastin said: Hèy John, it would be great to hear about how that Weiss LPS performs. Can I suggest to start another thread so this one doesnt drift much off topic. I understand the suggestion about a new thread. The following may fit in this thread though. After a few days of settling in the contribution of the of the PSU and the Rhein Z1 was totally seductive. The Rhein is 12V in and the power supply is switchable with a 6V setting. The Rhein and the Bonn N8 network switches are both made by Silent Angel. The two Bonn N8s that I have are 5V in and the therefore the 6V should be fine. I was told that the outcome maybe even better with the switches than the PSU with the Z1. So, I ordered another one as I did not want to compromise the outcome of the Z1. This is my first experience with separate LPS. I am now a fan. The Second PSU arrived in the post this morning. Plugged in to the two network switches and there was an immediate step up benefit but nothing too excitable. Well, tonight things have changed noticeably with the new PSU and the two switches. Initially it was a better bass with both the two PSUs. Over a few hours the mids and the highs start to come through and feature more. The bass is less prominent. I expect that I will experience a few more days of improvement. The SQ in my system now is just sublime and just so silky smooth. John Edited September 22, 2020 by Assisi 1
frednork Posted October 14, 2020 Posted October 14, 2020 an interesting review here https://forum.polkaudio.com/discussion/190575/etherregen-audiophile-ethernet-switch-video-and-audio-review shows ethernet cable differences, literally darker background Figure 17. Pixel level pictures, side by side. Left to right: Amazon Cable Matters Cat6a, Blue Jeans Certified Cat6a, AudioQuest Cinnamon Cat7, AudioQuest Vodka Cat7, Revelation Audio Labs Prophecy CryoSilver Cat8+.
Guest rmpfyf Posted October 15, 2020 Posted October 15, 2020 23 hours ago, frednork said: an interesting review here https://forum.polkaudio.com/discussion/190575/etherregen-audiophile-ethernet-switch-video-and-audio-review shows ethernet cable differences, literally darker background Figure 17. Pixel level pictures, side by side. Left to right: Amazon Cable Matters Cat6a, Blue Jeans Certified Cat6a, AudioQuest Cinnamon Cat7, AudioQuest Vodka Cat7, Revelation Audio Labs Prophecy CryoSilver Cat8+. Don't understand why a photo of a paused video is used to evaluate this. At 24FPS + whatever the television does there's many variables at play here. Might have been easier, 'frinstance, to create a high-def video of a still picture and to photograph that much; to have added some markers for framing and colour samples for colourimetry/calibration, then to plot the per-pixel differences to the source material. Interesting experiment but it needs control. Still... it's in the right direction IMHO.
RussB Posted October 15, 2020 Posted October 15, 2020 I’m being picky, but ideally he would have also taken the photos at a lower ISO than 1000, for less grain, and a smaller aperture than F3.2 such as F8 where that lens tends to be sharpest! But that would require a longer shutter speed which would be ok on his tripod unless the pause creates some minor movement of the screen image.
pwstereo Posted October 15, 2020 Posted October 15, 2020 59 minutes ago, RussB said: ...a longer shutter speed... This whole situation seems full of potential variation. So the OLED screen is refreshing at 60 times per second (or possibly 120 times per second), and the shutter opens and closes at random within that second for 1/100th of a second. This randomness of the sampling seems fraught with potential for variation. I'd like to see a set of ten images of the same cable/equipment combination side by side. A longer shutter speed, like 10 seconds would seem more sensible, to average the refresh of each screen pixel and minimise the potential for variation due to partial/complete refresh during the short shutter speed used. 1
frednork Posted October 16, 2020 Posted October 16, 2020 3 hours ago, rmpfyf said: Might have been easier, 'frinstance, to create a high-def video of a still picture and to photograph that much; to have added some markers for framing and colour samples for colourimetry/calibration, then to plot the per-pixel differences to the source material. 2 hours ago, RussB said: ’m being picky, but ideally he would have also taken the photos at a lower ISO than 1000, for less grain, and a smaller aperture than F3.2 such as F8 where that lens tends to be sharpest! But that would require a longer shutter speed which would be ok on his tripod unless the pause creates some minor movement of the screen image. 1 hour ago, pwstereo said: I'd like to see a set of ten images of the same cable/equipment combination side by side. A longer shutter speed, like 10 seconds would seem more sensible, to average the refresh of each screen pixel and minimise the potential for variation due to partial/complete refresh during the short shutter speed used. They all sound like excellent ideas. Would love to see someone confirm/disprove that result considering the paucity of data in this space. Unfortunately I am not very knowledgable in image reproduction and analysis.
MLXXX Posted October 16, 2020 Posted October 16, 2020 (edited) Underneath the above comparative images, the following words appear: "Figure 4. Left - picture with NAS and stock switch mode power supply. Right - picture with NAS and TeraDak DC-200VA linear power supply. There were drastic differences in detail, color saturation, and clarity between pictures with the NAS's stock switch mode power supply and replacement upgrade linear power supply." Drastic indeed! How a mere change in the power supply could be responsible for such dramatic differences is hard to fathom. Did the power supplies deliver different DC voltages? Something very odd going on there either with the equipment under test, or the experimental technique used to capture the images produced by the equipment under test. Edited October 16, 2020 by MLXXX 1 1
pwstereo Posted October 16, 2020 Posted October 16, 2020 11 minutes ago, MLXXX said: Underneath the above comparative images, the following words appear: "Figure 4. Left - picture with NAS and stock switch mode power supply. Right - picture with NAS and TeraDak DC-200VA linear power supply. There were drastic differences in detail, color saturation, and clarity... Yeah, that's a weird one. When I get home on Sunday afternoon, I might do a little bit of testing. I've got a similar LG OLED TV, the same camera (I'll use a macro lens that is better suited to this sort of work), and I can make a series of images at the same shutter speed, plus a series with what I think is a more sensible (long) shutter speed. There's more going on than just cable/component/psu changes I think. 1
frednork Posted October 16, 2020 Posted October 16, 2020 (edited) ignore Edited October 16, 2020 by frednork
frednork Posted October 18, 2020 Posted October 18, 2020 Same review and additional info on setup and how/why here https://audiophilestyle.com/forums/topic/58910-etherregen-formal-video-imaging-and-audio-review/#comments
pwstereo Posted October 18, 2020 Posted October 18, 2020 (edited) Hmmmm so, I set up my Nikon D800E with a Nikkor 200mm f/4 AF-D macro lens around 82 inches from the TV screen. TV is a 65 inch LG OLED similar to the one used in the test referenced above. I made a series of ten exposures for each camera/lens setting I tried. The interval between frames would range from 0.5 seconds to 3 seconds (I tried not to create a pattern or specific time interval, so that each frame would be unique in the amount of TV pixel refresh occurring). Image processing was Adobe Lightroom from the Nikon raw files, white balance adjusted to Shade for all frames, no other adjustments for the first eye image sets. Exported to jpg files 1200 x 800 pixels. My test image was a still photo of an eye, (I used two different ones, a brighter one and a darker one), loaded from a USB thumb drive plugged into the TV. No power cables or Ethernet cables were harmed during these tests. Here are the first set of ten images - 1/90 sec at f4.8 (similar settings to the original tester). Edited October 18, 2020 by pwstereo Spelling
pwstereo Posted October 18, 2020 Posted October 18, 2020 Here are the second set of ten images - 1.0 sec at f32 (so trying to have lots of TV pixel refreshes in each image).
pwstereo Posted October 18, 2020 Posted October 18, 2020 I'll comment at this stage that I see small and subtle variations from photo photo. Highlight detail varying (the tiny red veins around the coloured iris area), shadow-mid transitions varying (between the black and coloured patterns of the coloured iris area), and occasional variation in the blown white areas of reflected light at the bottom right inside the eyelid.
pwstereo Posted October 18, 2020 Posted October 18, 2020 Moving on to the second, darker coloured eye. Here are the third set of ten images - 1/90 sec at f4.2 (similar settings to the original tester). This set has had an exposure adjustment to normalise the tonality (all frames synch'd with same adjustment).
pwstereo Posted October 18, 2020 Posted October 18, 2020 I'll comment again at this stage that I see significant variations from photo photo. Colour varying (the green in the coloured iris area), shadow-mid transitions varying (between the black and coloured patterns of the coloured iris area), and occasional light tone variation in the lighter parts of the coloured iris patterns. To me, these variations are of a similar degree to those shown in the original test concerning cable changes.
pwstereo Posted October 18, 2020 Posted October 18, 2020 Here are the fourth set of ten images - 1.0 sec at f32 (once again trying to have many TV pixel refreshes in each image).
pwstereo Posted October 18, 2020 Posted October 18, 2020 And a final set using the second eye and a 1/30 sec exposure at f5.6 just to add a different data point, this would have had roughly 4 TV pixel refreshes per image. (The TV has a 120 Hz refresh rate.)
ikhuong Posted October 18, 2020 Posted October 18, 2020 Can you comment on each photo set which one is with etherregen?
Guest rmpfyf Posted October 18, 2020 Posted October 18, 2020 This all means nothing until you can control externalities. Put a timestamp on the frame. Better yet, have a video where there only frame is one still picture. Put some colour references you can calibrate to. Put some dimensional references you can zoom, crop and distort (e.g. pincushion) against. Black out the surrounding environment so the only light source in the room is the TV. When you're done with that and have applied any necessary adjustments, computing a difference plot is easy. When you have difference plots, computing absolute magnitudes are easy. Otherwise this lacks an amazing amount of structure.
dbastin Posted October 18, 2020 Author Posted October 18, 2020 Dudes ... this thread is about Ethernet Cables for Audio. I assume Frednork was initially highlighting an assessment that shows higher grade cables make a difference. Can we please try to highlight the relevance of this to audio. And not get into a lengthy discussion about photographic methodology. If you wish to try to demonstrate that regular Ethernet cables are adequate, please do so in Part B. Cheers 3
Guest rmpfyf Posted October 18, 2020 Posted October 18, 2020 3 hours ago, dbastin said: Dudes ... this thread is about Ethernet Cables for Audio. I assume Frednork was initially highlighting an assessment that shows higher grade cables make a difference. Can we please try to highlight the relevance of this to audio. And not get into a lengthy discussion about photographic methodology. If you wish to try to demonstrate that regular Ethernet cables are adequate, please do so in Part B. Cheers You could use the same means being proposed in the power cable test thread, y'know... might actually be super interesting as you could play audio from various points in a network chain and see about any differences. Jitter differences under sampling resolution might be difficult to replicate but if you showed anything up it'd be interesting. You could literally stream a sound file of one frequency and do a silly-long FFT to see if anything were going on. Could be pretty cool.
pwstereo Posted October 18, 2020 Posted October 18, 2020 4 hours ago, ikhuong said: Can you comment on each photo set which one is with etherregen? All photos are without any special items or equipment. That was the point of the test series. To show that the differences from frame to frame in the test pointed to above were likely to be shot to shot variation, not due to the changes in equipment or cables. There was no change at all between any of the images (apart from using two different images, one that had a lot of high values, and one that had more low values). 1
pwstereo Posted October 18, 2020 Posted October 18, 2020 (edited) 3 hours ago, rmpfyf said: ...only frame is one still picture. Put some colour references you can calibrate to. That's the whole point, nothing at all changed between frames. It is one image that is sized to the pixel dimensions of a 4k screen. No need, all images shot raw, equal correction and processing applied to all in each set. It's not about colour accuracy, it's about shot to shot variation with no change at all in environment and equipment. It was intended to show that the test referenced above was not a valid demonstration of cable and equipment changes. Edited October 18, 2020 by pwstereo Spelling 1
Recommended Posts