SingMai SM03 came to a dead end?

[latreche34]

We've discussed SingMai SM03 here before, Unfortunatly I have some bad news regarding this FPGA capture device SM03 and other similar products from the same company, I let you read the e-mail I got from Daniel:

Quote:

Hello Xxxxx,

I am very sorry, but we have made the decision to withdraw the SM03 (and other module products) from sale.

I would like to explain the reason behind our decision.

The SM03 uses proprietary designs and these are implemented in a FPGA (field programmable gate array). The SM03 was designed using an Altera FPGA. You may be aware that there is a now a worldwide shortage of electronic components just as demand as increased (e.g. the increase in electric car sales). Overnight we were not able to source the device from any franchised distributor and they were not even accepting back orders. No date was given as to any expected deliveries and that is still the situation 18 months later. We had some stock of FPGAs and we bought some from a China company. However the FPGAs we bought had an error in the device code which made us suspect they were not original parts. We could still buy these parts but the price is now more than what we sell the SM03 for, so that is not viable.

So we made the decision to redesign the SM03 to use a new FPGA company’s devices. To replicate the Altera part we had to use two devices which caused issues because we had to split the design. The new devices were missing some features that the Altera one had but we thought we could design around that.

We have managed to get an SM03 working, but testing a batch of the boards always gives one or two that fail (have noisy outputs caused by timing issues). Despite all our efforts (and help from the FPGA manufacturer) we have not been able to solve the problem. We don’t want to ship a ‘working’ part only for it to give problems later. There is no alternative device we can use that meets our performance/features/price requirements.

It is not just the FPGA, other components we have had to source from non-franchised distributors or use more expensive versions of. And we, of course, are not the only company facing these issues; just last month a company we know well who have been trading for 28 years have closed their business because of component supply issues.

I am so sorry to make this decision, but the last thing I want to do is supply a product that we are not certain of and proud to sell.

I am sorry for the long explanation, and truly sorry for any inconvenience this may cause you.



Kind regards,



Daniel.

[Hushpower]

That's no good, was thinking of getting one.

[BarryTheCrab]

The SM-03 is back in production and for sale.

[latreche34]

I totally forgot to update this thread, They are using a different FPGA with a different design, How good compared to the original design remains to be seen.

I'm not aware if it's in production and certainly not for sale yet.

[BarryTheCrab]

It is on their website, no mention of price.

[latreche34]

Because it is not available for sale yet according to Daniel himself.

[kitty666cats]

I had emailed Daniel a while ago back when we thought this project was dead, he told me to keep an eye out around October and, sure enough, emailed me when the SM03 was once again listed on their site.

Let's go over these advertised details, verbatim:

"SM03 is an NTSC or PAL video decoder, noise reducer and synchroniser designed to produce high quality video images from jittery, noisy and/or out of specification sources.

It is ideal for applications in the:

Improved viewing of legacy NTSC and PAL video sources such as Laserdisc.

Recovery of analogue tape and video disc video for archiving and transfer to modern media.

Very low artefact decoding and noise reduction for MPEG encoding pre-processing (reduces compressed bandwidth requirement by up to 20% compared with conventional decoders).

Broadcast quality decoding of composite video images.

Enhancement of poor quality video images."

(image of front and rear panels)

"SM03 Specifications

Video Inputs:

NTSC/PAL/SECAM/NTSC-443/PAL60 standards.

CVBS (BNC: ~1.0V into 75Ω).

Y/C (2-BNC: Y ~1.0V into 75Ω, C ~0.3V into 75Ω). A male S-VHS to BNC cable is provided with the SM03.

Dual 12-bit ADC, 27MHz sample rate.

4dB overhead margin.

Video Output:

525i/625i serial digital interface (SDI - SMPTE-259M). BNC: 800mV pk-pk.

DVI: 525p or 625p (3D motion adaptive video deinterlacer)

Video Decoder:

Motion adaptive, 7-tap 3D comb filter (frame, field, line, notch selected on a pixel by pixel basis).

Noise reduction:

Adaptive 7x7 median filter (reduce impulse and 'salt and pepper' noise).

Motion adaptive 3D recursive filter.

ynchronisation:

Full frame synchroniser.

Crystal based output sync pulse generator. Compliant, low jitter output under all input video conditions.

Contrast Enhancement :

Adaptive histogram equalisation.

Chroma:

Chroma edge enhancement (no ringing or overshoots).

Audio:

Stereo analogue audio input (48kHz sampling).

Audio embedded into SDI output.

Programmable audio delay for lipsync compensation.

Audio delay can track synchroniser delay.

Control:

Colour LCD display with encoder/switch control.

Full Proc Amp controls (luma and chroma gain, brightness and hue).

Dimensions:

165mm x 100mm x 42mm (1U rack compatible).

Power Supply:

12VDC (90V-260V AC-DC adaptor supplied)."

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Sounds pretty promising and enticing, I hope it ends up being an excellent option and perhaps offsets the inordinate prices of the AVToolBox, DataVideo, et al. Looking forward to seeing some of the first few folks who get their hands on one for testing.

I sent Daniel a follow-up email of a feature request that is both probably FAR too late to implement, and far too niche for most users to even be interested in:

I asked him if there was any possibility of making the DVI output *DVI-I* instead of DVI-D so that it can output both analog and digital (DVI-D and ALSO RGBHV + YPbPr over the DVI. Seen this on many devices. I've actually seen, on a VERY marginal few devices, non-standard DVI pinout where they even implement pins for composite, Y/C and even stereo audio! Extremely rare to see this, though... I think I've primarily seen a small handful of Kramer devices that have said non-standard DVI pinout).

Also asked if he could add a functionality of the DVI output (whether it's strictly DVI-D, or the preposterous selection of outputs I mentioned above) to also output a non-deinterlaced 525i / 625i signal. Would be a very niche use case for most, but I have some personal reasons/gear where that output option would be of interest and benefit to me.

tl;dr I hope the SM03 lives up to all its claims and ends up being a great asset to all the folks on this forum, LaserDisc enthusiasts, and many others!

[lordsmurf]

Quote:

Originally Posted by kitty666cats

I hope it ends up being an excellent option and perhaps offsets the inordinate prices of the AVToolBox, DataVideo, et al.

I highly, highly doubt it. What we have here is another SDI appliance (SDI capture card), with a few processing "bells and whistles" that were desired 20 years ago (before Avisynth matured).

The Laserdisc features may be useful for that crowd, not sure.

[latreche34]

For the purpose that card was made for, DVI has absolutely nothing to do with it, DVI is video only, If monitoring is what's needed, HDMI will be a better choice as it carries both audio and video.

The feature that is convenient is a USB port, But he already explained that licensing issues are the barrier to go the USB route, So he left that to other companies that already invested in that technology and have the licensing to use it like Magewell and BM. Besides having SDI out (which is the actual standard that governs analog video capture when the first D1 machine used it) is better for easy signal troubleshooting. Also having SDI and USB/PCie/thunderbolt as separate entities gives you more options and can be adapted into any OS, any port (USB3, USB4, thunderbolt, expansion card), any connector style (A, B, C). SDI to USB/thunderbolt adapters are small nowadays, computer powered, So they don't take space or require extra cables and power adapters.

[latreche34]

They've announced the price so I assume it's ready for order, But at a higher price than the previous version, £1095, I can only assume because of the higher prices of the new chips. and yes I've seen prices of parts and devices in other fields gone 5 times higher due to chips new pricing.

[NJRoadfan]

How about pictures of some actual hardware? A drawing that was made in Paint and a wall of specs. screams vaporware to me. The BoM cost for this isn't going to be too different from a RetroTINK-5X or other video scaler. I don't see this using a high end FPGA either as it isn't upscaling to HD video.

The USB licensing is a non-issue if you are using someone else's interface chips anyway.

[latreche34]

If USB licensing wasn't an issue they would have implemented it, I disagree on upscaling to HD for capturing, analog video is interlaced and has to be captured that way, In post you can do non real time processing if you would like.

Scalers were never made to capture video, they are designed for display purposes on modern flat panels, this is a fact.

Yes cost of such devices is high for the average Joe, the same with anything else, You can buy a $20 Harbor Freight drill or a $200 Makita drill, they both drill holes but they are not made for the same person.

[NJRoadfan]

Didn't say anything about the lack of scaling. I'd rather it just capture 480i video. My concerns were the lack of evidence of an actual product to sell and the high asking price.

For comparison's sake, the OSSC Pro scaler was finally announced for sale this past week. The unit had a similar delay in development due to FPGA chip shortages. During that time there were photos of actual working prototypes and units were seeded to testers during firmware development. It uses a pretty high end Cyclone V FPGA and the initial price is €290.00.