← Back to Home

Borescope 2026: diameter, depth of field, pixels

borescopeendoscopeinspection camera

People buy a borescope to answer one question: can I see the thing I cannot reach? The three numbers on the listing are not three independent features. Probe diameter decides which holes you can enter, depth of field decides at what distance the image is still sharp, and pixel count decides how small a defect you can separate. They multiply, they do not add. Zero one of them out and the other two stop mattering.

I actually opened seven live amazon.com product pages, checked the title, the bullet points and the spec table on each, and then pulled the spec sheets from teslong.com to do the arithmetic. I have not torn anything down and I have not put a scope in water. Every distance and pixel figure below is calculated, not measured with my hands, and that distinction matters.

This post contains Amazon affiliate links: if you buy through one, I earn a commission and the price you pay does not change. Prices were displayed in euros or Singapore dollars at capture time depending on region, so unless the brand store lists a USD price I write "check the listing". Data source: public web retrieval, as of 2026-11-04.

Why borescope spec sheets are harder to read than most gear

Most product specs sit side by side without touching: screen size, port count, weight. A borescope is different. Its three core numbers are tied together by the same optics and the same cost budget, so improving one always taxes another.

It gets worse because sellers know buyers scan titles. The title says one number, the bullets say a second, the spec table says a third. Seeing 180° and 210°, or 5 mm and 5.5 mm, or "1.0MP" and "1080P" on the same page for the same unit is not rare. This is not one bad seller. It was the recurring pattern across the pages I checked.

So the right order is not "which number is highest". It is: how narrow is my hole, how far away is the target, how small is the smallest defect I must resolve, and then find a model where all three hold at once.

⏳ TL;DR

🥇 You need to steer around a bend, into a cylinder or behind an elbow: Teslong TD500 6.5 mm dual-lens articulating model — 210° two-way articulation, 5" IPS screen, 32 GB card, 6.5 mm probe, rated IP67. The defect is in section seven: the official page states both a "1.0 MP sensor" and "1920x1080 photos", and those differ by 2.25x.

👉 Check the TD500 6.5 mm on Amazon >>

🔧 Straight viewing is enough, 1.5 m of reach, you want a screen: Teslong 5" IPS dual-lens model — 5 ft cable, 3500 mAh battery, front plus 90° side camera, official runtime "up to 6 hours". The defect: the title says 5 mm and the bullets say 5.5 mm.

👉 Check the Teslong 5 mm dual-lens on Amazon >>

📱 You already carry a screen: Teslong 10 ft USB-C model — plug in and shoot, 10 ft flexible cable, 8 LEDs, 1.0 MP. The defect: no battery, no screen, and it drains your phone.

👉 Check the Teslong 10 ft USB-C on Amazon >>

None of the three had a stable USD price at capture time; check each listing. If all you want is a look behind a hole you just drilled, skip to branch three of the decision tree.

Measure the hole first: diameter decides access, and also how close you can focus

Three diameters cover most of the market: 5.5 mm, 6.5 mm and 8.5 mm. US listings prefer inches, so you will also meet 0.21 in, 0.22 in and 0.25 in. Converted: 0.21 in = 5.33 mm, 0.22 in = 5.59 mm, 0.25 in = 6.35 mm. Two scopes advertised as "0.21 in" and "5.5 mm" may differ by 0.2 mm, or may be the same probe with two spellings. The Teslong NTS500 title says 0.21 in while a sibling model says 5 mm in the title and 5.5 mm in the bullets.

The hard constraint is obvious: if the probe is wider than the hole, everything else is void. What almost no listing tells you is the reverse cost — a thinner probe focuses farther away, not closer.

Teslong's official TD450S sheet (8.5 mm probe) gives depth of field as 0.79 to 3.94 in (20 to 100 mm) forward, and 0.4 to 2.76 in (10 to 70 mm) for the side camera. A 5.5 mm articulating model lists "optimal depth of field 3 to 8 cm" in its own spec block. Put both on one axis: the 8.5 mm probe focuses down to 20 mm, the 5.5 mm one stops at 30 mm.

The reason is optical. Shrink the probe and the lens aperture shrinks with it, so light throughput drops; to keep producing an image in the dark, the focus band gets pushed forward and narrowed. Thin buys access, not proximity. If your target sits 20 mm inside the opening — a crack along the bore wall, a joint just past a bend — the 8.5 mm unit images it more easily, provided the hole accepts 8.5 mm.

Depth of field is fixed; cable length is not

This is the expensive misunderstanding. Cable length is mechanical: 5 ft, 9.8 ft, 50 ft are all buildable. Focus distance is optical: once a scope is fixed-focus, that band is locked. The two are unrelated.

The Teslong MS450 is a 50 ft (15.24 m) drain scope. Its official spec block lists depth of field as 30 mm to 100 mm for both the front and side cameras, 70° field of view, 1.0 MP. Put that next to the length. The probe reaches 15 metres; the layer it can resolve is the 3 to 10 cm in front of the lens. To cover 15.24 m of pipe wall you would stop every 10 cm, roughly 152 stops. That is not a sweep, it is a crawl.

The same arithmetic explains why long-cable models always add "use with a drain snake". It is not about push force. The distance at which the image is legible is simply too short, and only repeated positioning compensates.

Field of view versus distance is the number no listing provides. Visible width W at distance d for field angle θ is W = 2·d·tan(θ/2). At 70°:

The same 70° lens shows a 28 mm patch when you hold it 2 cm away and 140 mm when you back off to 10 cm. "Wide angle sees more" is only true once you retreat — and retreating collapses pixel density at the same time.

By how much? Divide visible width by horizontal pixel count to get millimetres per pixel:

DistanceWidth at 70°At 1280 pxAt 1920 px
20 mm28.0 mm0.0219 mm/px0.0146 mm/px
30 mm42.0 mm0.0328 mm/px0.0219 mm/px
45 mm63.0 mm0.0492 mm/px0.0328 mm/px
100 mm140.0 mm0.1094 mm/px0.0729 mm/px

Here is the test: reliably identifying a 0.1 mm crack needs at least two pixels across it, so mm/px must be ≤ 0.05. On a 1280-pixel sensor that holds only within 45.7 mm. At 1920 pixels the limit moves to 68.6 mm. The MS450 at 100 mm gives 0.109 mm/px, so a 0.1 mm crack occupies one pixel and is indistinguishable from noise. A 50 ft probe gets you to the site. It does not promise you can read it.

The pixel ledger: 2.25x between 1920x1080 and 1.0 MP

Now the contradiction I kept meeting. 1920x1080 = 2,073,600 pixels, about 2.07 MP. 1280x720 = 921,600, about 0.92 MP. The gap is 2.25x.

Teslong's own TD500 page reads: "1.0 MP Sensor Captures 1920x1080 Resolution Photos & Videos". Literally that is not a lie — the file is 1920x1080. But the sensor is 1.0 MP, which is the 1280x720 class. The missing 2.25x is interpolated. Interpolation adds no detail; it softens edges. So the question to ask is whether you want a 1080P file or 1080P detail.

The same accounting problem hides in the screen, and it is worse. The TD450S official spec: 4.5-inch screen at 854x480, capture and video at 1920x1080. 854 × 480 = 409,920 pixels, one fifth (1/5.06) of 1080p. On site, the screen shows a fifth of the pixels in the file, so you cannot really judge focus or confirm you caught the crack. You pull the microSD card and look on a computer.

Some sellers do get this right. The MS450 description says its 4.5-inch 854x480 IPS panel is "over three times the clarity compared to most budget endoscopes with 4.3-inch 480x272 TN screens". That claim is checkable: 480 × 272 = 130,560, and 409,920 ÷ 130,560 = 3.14x. The claim holds. Marketing that survives arithmetic is worth more than the word "HD".

The decision tree: five branches, one model each, plus a fallback

Ask yourself three questions in order, then take the branch.

Branch one: opening under 6 mm, and you must see sideways or into a cylinder. Take a 5.5 mm articulating model such as the Teslong 5.5 mm two-way articulating borescope. Its spec block reads: Φ5.5 mm probe, 1.5 m cable, optimal depth of field 3 to 8 cm, 75° field of view, 1280x720 screen, 720p video. Defects in section seven — its title and its spec block disagree on articulation angle.

Branch two: opening 6.5 mm or larger, and you need to work around an elbow or along a cylinder wall. Take the TD500 6.5 mm dual-lens, B0CHW4FLS7. Official figures: two-way 210° articulation, 5.08 ft (1.55 m) probe, 6.5 mm dual lens (front plus side), 5-inch screen, 3000 mAh, 4 hours runtime, IP67, 32 GB card included. The 210° matters because it can hook back 30° past perpendicular — the move a 180° unit cannot make.

Branch three: straight viewing only, target within 1.5 m, and you do not want to pay for articulation. Take the Teslong 5 mm dual-lens with screen: 5-inch screen, 5 ft cable, 3500 mAh, front plus 90° side camera, 8 LEDs, 32 GB card. Wall cavities, finding a break inside conduit, checking cable routing behind a case — this tier covers it. It is also the one pick of the three with a 3500 mAh cell.

Branch four: pipe runs longer than 3 m. Take the Teslong MS450 50FT drain model: 15 m replaceable probe, 8 mm diameter, 1.0 MP, 70° field, 30-100 mm depth of field, 4.5-inch 854x480 screen, 2500 mAh. Read the 152 stops paragraph before ordering. These products deliver arrival, not legibility.

Branch five: your phone is the screen. Take the Teslong 10 ft USB-C model: 10 ft (3.05 m) flexible cable, 8 LEDs, 1.0 MP, IP67, works with iPhone 15/16/17 and Android. No battery, no screen, no card — the cost sits entirely in the cable and the lens.

Fallback when unsure: pick the listing that is written consistently, not the one written loudly. The Teslong NTS500 states "5-inch 1280x720" and "720p video" in the same breath, no hedging. That consistency is itself a quality signal. The cost is that it is 720p, and the price displayed in euros at capture time.

One control sample worth naming: Klein Tools ET19 (B0DLDS8N7D), 5.5 mm probe, 9.8 ft (2.99 m) armored snake, 4.3-inch screen. Its spec field says "Video Capture Resolution 1080p" while the official description on the same page says "720p resolution". Two answers, one page. I am not dismissing it; I am pointing out that this kind of field conflict is not an outlier.

The real defects on seven models in stock

Teslong TD500 6.5 mm (B0CHW4FLS7)

Teslong 5 mm dual lens (B09Y1F2MPB)

Teslong 5.5 mm two-way articulating (B0DT4H7QQM)

Teslong MS450 50FT (B08QCMT89N)

Teslong 10 ft USB-C (B0C1GFBTDC)

Teslong NTS500 (B07SQ4FLYX)

Klein Tools ET19 (B0DLDS8N7D, control sample)

Three expectations that cost people money

Expectation one: IP67 means the whole unit can go underwater. The full wording is "maximum depth of 1 meter up to 30 minutes under IEC standard 60529", but the Teslong TD450S page immediately adds the qualifier: "The IP67 rating applies to the camera probe only." Klein phrases it as IP67 for camera head and cable. Handles, screens and buttons are outside the promise. Do not dunk the unit.

Expectation two: a longer cable sees farther. See the depth-of-field section. Length is travel, not imaging.

Expectation three: more pixels is better. Going from 1.0 MP to 2.0 MP only pays off if you can hold the probe steady inside 3 to 10 cm. Five millimetres of hand shake at 45 mm is a 10% frame shift, which dwarfs any pixel difference.

FAQ

Q: If a borescope says IP67, can I drop it into a drain?

A: Read the qualifier. Under IEC 60529, IP67 is dust-tight (6) plus water ingress protection (7) at 1 m for 30 minutes, but makers routinely limit it to the camera probe, or to probe and cable. The TD450S page states IP67 covers the probe only; the NTS500 page states no limit at all. Confirm the scope of the rating on the listing rather than inferring it from the model number.

Q: Should I compare 1080P or 1.0 MP?

A: Decide whether you want a file or detail. Output resolution describes the file; sensor pixels cap the detail. 1920x1080 is 2.07 MP, 1.0 MP is in the 0.92 MP class, a 2.25x gap; when a listing shows both, the output is interpolated. For documentation, the 1080P file is convenient. For judging 0.1 mm defects, go back to depth of field and working distance — those decide it.

Q: Can a 50 ft probe see the far wall of a 15 m pipe?

A: Not by sweeping. The MS450's official depth of field is 30-100 mm, so only the 3-10 cm layer in front of the lens is sharp. Covering 15.24 m means roughly 152 stops, capturing at each one. The product category is about arrival; reading comes from the frame-by-frame record.

Q: How much does 180° versus 210° articulation matter?

A: It is the difference between reaching perpendicular and hooking back. 180° takes the tip from straight to square; 210° adds 30° past square, which is what you need behind an elbow or along a cylinder wall. Note that some listings say 180° in the title and 210° in the specs — trust the spec block and the official sheet.

Q: Screen built in, or phone-based?

A: Ask whether you have a second hand free. Screen models (3000-3500 mAh) work one-handed and leave your phone alone; phone models are cheaper and lighter but drain the phone and are awkward to hold over a deep hole. Bench work favours a screen; pocket carry favours the phone version.

Closing

The three numbers are not a leaderboard, they are simultaneous conditions: the opening caps diameter, the working distance sets the focus window, the smallest defect sets pixel density. Only a model where all three hold is worth buying, and the one with the biggest numbers is not automatically it.

If you keep one line, keep this: diameter buys access, depth of field buys legibility, resolution buys separation, and none of them substitutes for another. For the same style of spec reading, see working distance and magnification limits on repair microscopes and NETD and resolution on thermal cameras.

👉 Join Xiaomi MiMo Platform: Leading AI model platform with cost-effective inference

👉 Join Aliyun AI: Top AI products with exclusive coupons for business innovation

📌 This article was AI-assisted generated and human-reviewed | TechPassive — An AI-driven content testing site focused on real tool reviews

🔗 Recommended Tools

These are carefully selected tools. Using our affiliate links supports us to keep producing quality content:

As an Amazon Associate, I earn from qualifying purchases at no extra cost to you.

☁️ DigitalOcean Cloud ⚡ Vultr VPS 🤖 QoderWork CN (Refer & Earn) ☁️ Aliyun AI Products 📚 WordPress Books 🧠 AI & LLM Books 🏗️ System Design Books 🐳 Docker & K8s Books 🐧 Linux Admin Books 🐍 Python Books 🗄️ Database Books 🚀 DevOps & SRE Books 🤖 Xiaomi MiMo Platform
← Back to Home